This repository contains code for performing image classification on the CIFAR-10 dataset using Convolutional Neural Networks (CNNs) implemented in PyTorch. The code is inspired by a Kaggle post by Shadab Hussain titled "CIFAR 10 - CNN using PyTorch" and has been adapted to a Jupyter Notebook format with CUDA acceleration.
The CIFAR-10 dataset consists of 60,000 32x32 color images in 10 classes, with 6,000 images per class. The task is to classify these images into one of the following classes: airplane, automobile, bird, cat, deer, dog, frog, horse, ship, and truck.
To run the code in this repository, you'll need:
- Python 3.x
- PyTorch
- Jupyter Notebook
- CUDA-enabled GPU (optional, but recommended for faster training)
# 1. Clone this repository
git clone https://github.com/noobsiecoder/cifar-10-image-classification.git
cd cifar-10
# 2. Install the required dependencies
pip install -r requirements.txt
# Open the Jupyter Notebook
jupyter notebook main.ipynb- The original inspiration for this code comes from Shadab Hussain's Kaggle post on CIFAR 10 - CNN using PyTorch.
- The CIFAR-10 dataset is provided by the Canadian Institute for Advanced Research (CIFAR).
This code is provided under the MIT License.